The present invention relates to the field of audio reproduction systems and, more particularly, to high fidelity phonograph preamplifier networks.
The normally quoted specifications of a phonograph preamplifier network, such as frequency response, noise and distortion, are of limited utility in determining how a preamplifier will sound to a listener. This is due to a variety of factors, including (1 ) to inability of conventional test measurements to measure performance under transient conditions, (2 ) the failure of performance test results to be weighted for subjective human perception or annoyance values, and (3 ) the effect of out-of-the-audio-passband signals on in-band performance due to inter-modulation in the preamplifier or other portions of the system.
Typically, analytical performance measuring techniques are based on gross simplifications of the actual conditions under which a preamplifier operates. Accordingly, the resultant measurements bear a relatively low correlation with the critical listening experience.
Often, subjective listening test comparisons are the most effective techniques of evaluating preamplifier performance. Of course, in such comparisons, normal use conditions must be established for the systems under test, e.g. the cartridge must be properly loaded, both resistively and capacitively, for each preamplifier under test, and preamplifier gains of the comparison tests must be very accurately matched, often to within a fraction of a decibel. In addition, grounding and shielding in test configurations must be done with great care since the preamplifier outputs must be physically close to the inputs in order to permit switching back and forth for comparative listening. Under such conditions, double blind listening tests have proven to be quite effective in providing comparative evaluations of preamplifier networks.
Such listener tests demonstrate that an important area of preamplifier performance is the infrasonic response, i.e. in the 1- 20 Hz range. At such frequencies, there are substantial detectable effects due to the driving of loudspeaker systems well below their nominal cutoff frequency, i.e. where the cone is essentially unloaded (especially with vented-box loudspeakers), and due to the cartridge-caused overload and intermodulation in tape machines and power amplifiers.
A particular source of the infrasonic response problem is the characteristic peaking of warp-caused distortion in phonograph records at frequencies on the order of 4 Hz. With typical tone arms and typical cartridges used in high fidelity phonographs, the tone arm and cartridge form a resonant system with a moderately high Q and a peak at frequencies between 4 and 8 Hz. As a result, there appears in many systems a relatively high degree of infrasonic noise at frequencies in this range. This problem has been recognized to some extent in the prior art with the typical approach solution being the selective damping of the tone-arm-cartridge resonant system. Systems utilizing this relatively limited effectiveness solution fall short of desired performance characteristics.
A further problem with the prior art systems is based on the high frequency interaction of the preamplifier input impedance with the source impedance provided by the cartridge. It has recently been recognized in this context that the cartridge impedance includes a generally overlooked component, i.e. the sum of frequency-dependent resistive losses due to eddy current and hysteresis losses associated with coil in the cartridge. This high frequency interaction problem was not significant in vacuum preamplifiers, due to their characteristic input capacitance and resistance, and consequently, there has been little if any preamplifier development directed to eliminating interaction of the cartridge source impedance with that of the preamplifier input stages and interconnecting cable. Typically, cartridge specifications state the "proper" load to insure response to the specifications, yet rarely if ever are the impedances of cables or preamplifier inputs specified completely. The so-called proper load is defined in terms of fixed resistive and capacitive values, for example, a cartridge terminating resistance is specified to be 47 Kohm at the input to a preamplifier buffer. Comparative listening tests indicate that such a termination does provide satisfactory RIAA equalization. However, in the system environment, there may be a substantial capacitive effect at the preamplifier input. With such a resistive and capacitive termination, there is a substantial interaction effect with the cartridge. The failure to effectively consider the frequency dependencies of the cartridge source impedance result in significant degradation factor in the prior art phonograph systems.
Accordingly, it is an object of the present invention to provide a high fidelity phonograph preamplifier network characterized by a substantial infrasonic cutoff response.
Another object of the present invention is to provide a high fidelity phonograph preamplifier network characterized by relatively low high frequency interaction between the preamplifier input and cartridge.